comparing ablation methods
Whether you're tracking coins, torches, someone's ability to keep marching, their reserves of magical power, research progress, or the chance that a relationship will fray with neglect or that monsters will appear when you make noise, there are a bunch of ways you can track "how many of (chipping-awayn-ness) until (thing happens)." These can be mathematically equal in terms of initial expected value (for instance, 20 spaces in a progress bar = Ud12 = 5% flat chance each time) but differ significantly in terms of where risk is concentrated and when information becomes available.
Cardinal increments: There is a flat number that is ablatively chipped away, and is known from start to finish. This is how dnd has classically thought of its main resources (hp, gold, torches, rations, encumbrance, charges, spell slots) other than time; relatively modern-seeming procedures like BitD clocks and Ironsworn's progress bars are also basically just this.
If you know how much of something there is, and that's a concrete number of objects - coins, rations, torches - cardinal increments tracks both the world and your character's understanding of that world (assuming they can count it.)
I find hp a bit dissatisfying for because the things it's tracking - whether that's "meat points" or endurance, positioning, etc - doesn't really fall into that basket, and because it back-concentrates uncertainty.1 I find the Underclock dissatisfying for similar reasons.
Flat odds: here, each time chipping-away happens, there's a flat chance that the thing happens. Random encounters are the most classical example.
Gambler's fallacy but real: here, each time chipping-away happens, there's a chance that the thing happens, and if not, that chance is higher next time. The one put to most general use is the usage die,2 but we also see this in, e.g., durf's damage system.
The usage die seems to be most often used to keep track of "boring equipment," like torches and rations, on the grounds that it reduces bookkeeping. This is a noble goal in itself, but I find it a poor fit for that kind of thing: if I have Ud6 of a resources, I have somewhere between 2 and infinity uses left, with an average of 5. There's no way when I'm looking through my quiver to count arrows that I end up with that level of uncertainty!3 But if I'm in the middle of a long forced march, wondering how many more hours I can stay awake and keep trekking, it's just about perfect.
Lindy. This is the inverse of GFBR - with percentage of the thing happening going down each time - and I don't know how consciously it's designed around but it does arise naturally whenever there's a flat chance "in reality" but players don't get access to what that chance is. If the GM chooses some percentage of the thing happening per unit of time or per triggering condition, each time it doesn't happen, that's evidence that the real chance is lower than you previously thought.
my preferences
In case it's not entirely obvious, my preference is for:
- Cardinal increments for concrete resources, such as rations, money, or ammo
- GFBR for resources that "deplete" but with only vibe-y indications of how far along they are, such as bodily energy, the progression of disease, or the goodwill of allies
- flat percentage for indifferent events in the world (which becomes Lindy if you/whoever you're pretending to be don't have a good guess about the percentage)
- if there would be a nonzero chance of the thing happening, there should be a nonzero chance of the thing happening4
Depending on how frustrated I get with Google Sheets, the logical next move is comparing a few concrete proposals, with easy ways of converting between them.
as articulated by Orbital Crypt in Weapons Kill People: "why is it, other than inertia, that a dagger deals 1-4 hit points of damage, and my fighting-man has 10 hit points, making it textually impossible for me to be stabbed to death until an arbitrary number of successful hits? Why is the (on average) fourth hit the one that actually kills me dead? Why do serpent venoms and pit traps only ever impair or exhaust, without actually killing? I'm sure you've all had at least one conversation about this in the past, in which someone (maybe even YOU!) explained the abstracted nature of TTRPG combat, how gameplay comes from resource management and tactical decision making, yadda yadda. The simple question stands: Why how come the bullet can't won't not kill you?"↩
usage die is also a bit cardinal-increment-y compared to the most classic GFBR, and cardinal increments like hp and Underclock where depletion proceeds by random amounts are more like GFBR than when they always increment by a known amount.↩
usage dice do have the very nice property of quantum collapse math: if you roll the usage die you can have that many uses left and go down a die size, or most sensibly in the case of counting how much stuff you have, rolling each of them in turn and adding them up.↩
"would implies should" is a nice simulationist slogan, enough so that I'm sure it's been formulated before?↩